Use of antiviral drugs to treat HIV infection are often limited by the inability of the drug to access sites of infection or by drug toxicity. We are investigating the roles of specific xenobiotic transporters in the inability of HIV-1 protease inhibitors to penetrate the blood-brain barrier and in the renal toxicity of nucleoside phosphonates. To this end, we measure xenobiotic transport using confocal microscopy (isolated rat and mouse choroid plexus, rat and mouse brain capillaries and killifish renal proximal tubules) and conventional radiotracer uptake techniques (renal cortical slices and choroid plexus from rats and mice). With regard to poor brain penetration of HIV protease inhibitors, our experiments with isolated brain capillaries indicate that the ATP-driven drug export pumps, p-glycoprotein and Mrp2, are important components of the blood-brain barrier and that both ritonavir and saquinavir interact with these transporters. Two methods of circumventing the barrier have been successful in our in vitro and in in vivo experiments: specific inhibition of p-glycoprotein using PSC833 and transient opening of brain capillary tight junctions with alkyl glycerols. With regard to nucleoside phosphonate nephrotoxicity, both adefovir and cidofovir have been shown to be substrates for the organic anion transporter, OAT1, which is expressed in renal proximal tubule. Moreover, in model systems, nucleoside phosphonate toxicity appears to be positively correlated with cellular accumulation and with OAT1 activity. We have shown in intact renal proximal tubules that adefovir and cidofovir inhibit uptake of organic anions at the basolateral membrane mediated OAT1, but they also inhibit efflux at the luminal membrane mediated by the ATP-driven drug efflux pump, Mrp2. Since these drugs do not reduce transport mediated by basolateral Oat3 and luminal p-glycoprotein, they appear to interact directly with OAT1 and Mrp2 rather than inhibit metabolism. These results suggest that nucleoside phosphonates are substrates for Mrps as well as for OAT1; they raise the questions of how altered Oat1/Oat3 and Mrp2 function might influence nephrotoxicity and how nucleoside phosphonate accumulation in renal cells might affect Mrp2 function. Finally, the results provide a conceptual basis for understanding why nucleoside phosphonates do not accumulate in choroid plexus; this tissue expresses Oat3 not Oat1.